Exascale Hybrid Numerical-AI Ensembles for Operational Flood-Season Forecasting in East Asia: 15-km Decadal Hindcasts and 1-km High-Resolution Capability

2026-05-24

AI Summary

This study addresses key challenges in subseasonal-to-seasonal prediction of East Asian summer precipitation, including the spring predictability barrier, weak large-scale precursors, and highly nonlinear local convective extremes. To overcome these limitations, the work proposes a hybrid seasonal forecasting system that synergistically integrates kilometer-scale coupled regional numerical models with data-driven artificial intelligence. A large ensemble is generated through multiple physical parameterizations, initial perturbations, and stochastic physics, enabling the first-ever convection-permitting (kilometer-scale) numerical–AI integrated forecasts. Leveraging the LineShine supercomputing platform, the system completes a ten-year reforecast spanning 2016–2025, comprising 1,774 ensemble members, within 14.6 hours. The approach improves the prediction skill score from ECMWF’s 71.8 to 75.9, significantly enhancing operational forecast accuracy and lead time, while also supporting high-resolution typhoon simulations at weekly scales.

Abstract

Seasonal forecasting of summer rainfall in East Asia remains a grand challenge, as predictability at 3 to 6 month lead times is constrained by the spring predictability barrier, weak large-scale signals, and localized nonlinear convective extremes. We address this challenge with CAPES, which integrates a kilometer-resolution coupled regional model with atmosphere, land, and ocean components and a data-driven AI seasonal forecasting system. At 15 km resolution, the fused workflow combines 174 numerical members from varying start times, physics schemes, and parameter perturbations with 1,600 AI members generated from initial and physical perturbations. Using the full LineShine system, CAPES completes ten annual 1,774-member hindcasts for 2016 to 2025 within 14.6 hours, improving the mean prediction score from ECMWF's 71.8 to 75.9 and delivering a major gain in operational forecasting capability. The 1-km configuration further enables fine-scale typhoon simulation and establishes the feasibility of kilometer-scale fused ensemble forecasting on a one-week timescale.

Problem

• seasonal forecasting
• East Asia
• summer rainfall
• predictability barrier
• convective extremes

Innovation

• hybrid numerical-AI ensemble
• kilometer-scale forecasting
• seasonal prediction
• exascale computing
• flood-season forecasting